Back to EveryPatent.com
United States Patent |
6,068,028
|
De Ro
,   et al.
|
May 30, 2000
|
Yarn scanning process and yarn unwinding sensor
Abstract
In a method of scanning, with the aid of a sensor device, a yarn of a
predetermined length which is intermittently withdrawn from a winding
reservoir provided on the storage drum of a yarn feeding device for
weaving machines, the yarn pulse acceptance for exclusively at least one
first yarn pulse is changed at an increasing yarn speed and/or upon
generation of at least one first or each winding signal to a yarn pulse
acceptance for further faster yarn pulses and non-acceptance of
interference pulses that are slower than the second yarn pulses. A yarn
withdrawal sensor which is suited for said method is characterized in that
a filtering device is provided with two different selective filtering
modes that differ from each other by their acceptance of yarn pulses
generated at different yarn withdrawal speeds, and that the filtering
device is switchable at an increasing yarn withdrawal speed from a first
filtering mode to at least one further filtering mode.
Inventors:
|
De Ro; Ignace (Ypres, BE);
Lilja; Henrik (Angered, SE)
|
Assignee:
|
Iro AB (Ulrichehamn, SE);
Picanol N.V. (Ypres, BE)
|
Appl. No.:
|
983365 |
Filed:
|
May 18, 1998 |
PCT Filed:
|
July 18, 1996
|
PCT NO:
|
PCT/EP96/03177
|
371 Date:
|
May 18, 1998
|
102(e) Date:
|
May 18, 1998
|
PCT PUB.NO.:
|
WO97/04151 |
PCT PUB. Date:
|
February 6, 1997 |
Foreign Application Priority Data
| Jul 18, 1995[DE] | 195 26 216 |
Current U.S. Class: |
139/452; 242/364.8; 700/144 |
Intern'l Class: |
D03D 047/36 |
Field of Search: |
242/364.8
364/470.15
324/71.1
139/452
|
References Cited
U.S. Patent Documents
4407336 | Oct., 1983 | Steiner | 139/452.
|
4768565 | Sep., 1988 | Tholander | 139/452.
|
4848417 | Jul., 1989 | Dekker et al. | 139/452.
|
4877064 | Oct., 1989 | Pezzoli | 139/452.
|
5613528 | Mar., 1997 | Zenoni et al. | 139/452.
|
Foreign Patent Documents |
889 255 | Dec., 1981 | BE.
| |
0 176 987 | Apr., 1986 | EP.
| |
0 286 584 | Oct., 1988 | EP.
| |
647 999 | Feb., 1985 | CH.
| |
Primary Examiner: Falik; Andy
Attorney, Agent or Firm: Flynn, Thiel, Boutell & Tanis, P.C.
Claims
We claim:
1. A method of scanning a yarn of predetermined length which is
intermittently withdrawn during insertion cycles of a weaving machine from
a winding reservoir provided on a storage drum of a weft-yarn feeding
device, said weft-yarn feeding device including a withdrawal sensor which
produces yarn pulses wherein at least one said yarn pulse is produced
during passage of the yarn within one insertion cycle, said withdrawal
sensor further producing interference pulses due to passing particles
including dirt, said weft-yarn feeding device including a circuit wherein
a winding signal is derived respectively from said yarn pulse and is
transmitted to a signal-processing device, the method comprising the steps
of:
providing a band-pass filter assembly in said circuit which said band-pass
filter assembly has a yarn pulse acceptance which permits acceptance of
said yarn pulses which are relatively slow and weak;
accepting at least a first one of said yarn pulses with said band-pass
filter assembly; and
changing said yarn pulse acceptance of said band-pass filter assembly with
an increasing yarn speed and/or upon generation of at least a first said
winding signal which corresponds to said first yarn pulse; and
said yarn pulse acceptance after said changing permitting acceptance by
said band-pass filter assembly of further said yarn pulses which are
relatively fast and strong and preventing acceptance of said interference
pulses which are relatively slower or weaker in comparison with said
further yarn pulses to suppress false winding signals caused by said
interference pulses.
2. The method according to claim 1, wherein said band-pass filter assembly
has first and second filtering modes, said changing step including
upshifting said band-pass filter assembly from said first filtering mode
to said second filtering mode at an increasing yarn speed, said band-pass
filter assembly when in said first filtering mode accepting at least said
first yarn pulse which is relatively slow and weak and when in said second
filtering mode accepting said further yarn pulses which are relatively
fast and strong, said second filtering mode being predetermined such that
said interference pulses which are slower or weaker are filtered out with
respect to said further yarn pulses which are faster or stronger.
3. The method according to claim 2, further comprising the steps of
adjusting the yarn pulse acceptance prior to the occurrence of each said
winding signal, at least during an initial acceleration phase of said
insertion cycle, so that said band-pass filter assembly accepts said yarn
pulses which are relatively weak, and thereafter adjusting the yarn pulse
acceptance upon the occurrence of said winding signal to again accept said
yarn pulses which are stronger and faster and prevent acceptance of said
interference pulses.
4. The method according to claim 3, further comprising the step of
maintaining said yarn pulse acceptance for faster and stronger yarn pulses
for the duration of a time window which is shorter than a shortest time
period between two said winding signals of said insertion cycle which
occur successively.
5. The method according to claim 4, further comprising the steps of
supplying an upshifting signal to said band-pass filter assembly in
response to at least said first winding signal, said upshifting signal
being maintained for the duration of said time window.
6. In a withdrawal sensor for a weft-yarn feeding device which said
weft-yarn feeding device comprises a storage drum for a winding reservoir
and is used for intermittently feeding yarn of an adjusted yarn length to
a weaving machine during insertion cycles, said withdrawal sensor
comprising at least one receiver which during each said insertion cycle is
responsive to passage of said yarn to generate yarn pulses and also
generates interference pulses in response to passing particles, said
withdrawal sensor further comprising a circuit which is assigned to said
receiver and in which winding signals can be produced from said yarn
pulses, and a device which is connected to said withdrawal sensor for
processing said winding signals, comprising the improvement wherein said
circuit comprises a filter assembly having at least first and second
selective filtering modes which differ with respect to whether said filter
assembly accepts said yarn pulses which are strong or weak, said filter
assembly including a switching device which switches said filter assembly
from said first filtering mode to said second filtering mode due to an
increasing yarn withdrawal speed or after at least a first detected yarn
passage, said filter assembly when in said first filtering mode accepting
at least a first one of said yarn pulses which is relatively slow and
weak, and said filter assembly when in said second filtering mode
accepting said yarn pulses, which are relatively fast and strong while not
accepting said interference pulses.
7. The withdrawal sensor according to claim 6, wherein said filter assembly
comprises a band-pass filter assembly which prevents acceptance of said
interference pulses when in said second filtering mode, said interference
pulses being relatively slow or weak in comparison to said fast and strong
yarn pulses which are accepted by said band-pass filter assembly when in
said second filtering mode, said band-pass filter assembly accepting any
of said yarn pulses in either of said first and second filtering modes
which are below a predetermined upper yarn speed limit.
8. The withdrawal sensor according to claim 7, wherein said device
comprises a microprocessor and said band-pass filter assembly is connected
to said microprocessor which is fed with said winding signals, an
upshifting signal being in a stand-by condition in said microprocessor and
said upshifting signal being transmitted by said microprocessor to said
band-pass filter assembly after receipt of at least a first said winding
signal or each said winding signal.
9. The withdrawal sensor according to claim 7, wherein said filter assembly
is switched with each said winding signal, at least within an initial
acceleration phase of said insertion cycle, with an upshifting signal from
said first filtering mode wherein said yarn pulses which are relatively
slow and weak are accepted to said second filtering mode wherein said yarn
pulses which are relatively fast and strong are accepted and interference
pulses are not accepted, said filter assembly being held in said second
filtering mode for a duration of a time window, and an adjustable timing
or counting member being provided to define said time window which said
timing or counting member operates upon generation of said winding signal.
10. The withdrawal sensor according to claim 6, wherein said circuit is an
active amplifier and band-pass filter assembly.
11. The withdrawal sensor according to claim 7, wherein said band-pass
filter assembly is provided with a high-pass filtering mode and a low-pass
filtering mode of which said low-pass filtering mode can be disabled by an
upshifting signal, said band-pass filter assembly including resistors
which are arranged in parallel and connected to analog circuit components
and whose resistance characteristics are controlled by applying said
upshifting signal to said analog circuit components such that only said
high-pass filtering mode is operative when said low-pass filtering mode is
disabled.
12. The withdrawal sensor according to claim 7, wherein said band-pass
filter assembly includes upper and lower passage frequencies that which by
said filter assembly, said band-pass filter assembly having means for
elevating said lower passage frequency by an upshifting signal from a
predetermined basic value to a predetermined maximum value, said basic
value corresponding to a yarn speed of about 2 m/s for said first
filtering mode and said maximum value corresponding to a yarn speed of
about 10 m/s for said second filtering mode, said upper passage frequency
being respectively at a frequency corresponding to a yarn speed of about
120 m/s.
13. The withdrawal sensor according to claim 7, wherein said switching
device is connected to said circuit such that said band-pass filter
assembly is reset by said switching device into said first filtering mode
upon a standstill of said yarn or expiration of a time period defined by
said circuit.
14. The withdrawal sensor according to claim 7, wherein said band-pass
filter assembly comprises frequency band filters having different high and
low cut-off frequency settings, said frequency band filters corresponding
to said first and second filtering modes respectively wherein said
switching device switches between said frequency band filters, said
switching device including means for operating said switching device in
response to a yarn withdrawal speed or generation of at least said first
winding signal or each said winding signal.
15. The withdrawal sensor according to claim 7, wherein said circuit
comprises an adjusting device for adjusting a scanning sensitivity of said
withdrawal sensor which is dependent on yarn quality, said adjusting
device being uncoupled from said band-pass filter assembly by virtually
grounded analog circuit components for separately feeding sensitivity and
upshifting signal levels.
16. In a weft-yarn feeding device which said weft-yarn feeding device
comprises a storage drum for a winding reservoir and is used for
intermittently feeding yarn of an adjusted yarn length to a weaving
machine during insertion cycles, said weft-yarn feeding device including a
withdrawal sensor comprising at least one receiver which during each said
insertion cycle is responsive to passage of said yarn to generate yarn
pulses and also generates interference pulses in response to passing
particles, said withdrawal sensor further comprising a circuit which is
assigned to said receiver and in which winding signals can be produced
from said yarn pulses, and a device which is connected to said withdrawal
sensor for processing said winding signals, comprising the improvement
wherein said circuit comprises a filter assembly having at least first and
second selective filtering modes which differ with respect to whether said
filter assembly accepts said yarn pulses which are strong or weak, said
filter assembly including a switching device which switches said filter
assembly from said first filtering mode to said second filtering mode due
to an increasing yarn withdrawal speed or after at least a first detected
yarn passage, said filter assembly when in said first filtering mode
accepting at least a first one of said yarn pulses which is relatively
slow and weak, and said filter assembly when in said second filtering mode
accepting said yarn pulses, which are relatively fast and strong, while
not accepting said interference pulses.
17. The withdrawal sensor according to claim 16, wherein said weft-yarn
feeding device includes a stop device which is assigned to said storage
drum and is adapted to be moved back and forth between a stop position and
a passive position for said yarn for defining in said yarn feeding device
the predetermined yarn length for each said insertion cycle, said receiver
being arranged a short distance from said stop device in the direction of
motion of said yarn during withdrawal, said receiver being connected via
said circuit to at least one control device of said stop device.
18. The withdrawal sensor according to claim 17, wherein two said
withdrawal sensors are provided, one of said withdrawal sensors being
positioned a short distance in front of a stop element of said stop device
in a direction of motion of said yarn and another of said withdrawal
sensors being a short distance behind said stop element.
19. The withdrawal sensor according to claim 17, wherein said receiver is
axially offset relative to a stop element of said stop device in an axial
direction of said storage drum.
Description
FIELD OF THE INVENTION
The present invention relates to a method and to a yarn withdrawal sensor
therefor which accommodates interference pulses from a yarn sensor
arrangement caused by dirt, lint or the like.
BACKGROUND OF THE RELATED ART
In a method which is described in CH-B-647 999, winding signals are
produced from yarn pulses and are then counted. In practice, the correct
number of winding signals sometimes represents a number of withdrawn
windings that is either too great or too small, resulting in picks that
are too short or too long. For it sometimes happens that a free lint
cluster or a yarn component hanging onto the yarn (e.g. in multifilament
yarns) is dragged behind a withdrawn yarn winding and below the withdrawal
sensor and therebeyond, so that the sensor reports the cluster or
component as an additional winding. By contrast, when two adjacent
windings are withdrawn in close vicinity below the winding sensor, only
one winding signal is produced for the two windings.
EP0 286 584 B1 discloses another method of this kind, in which yarn pulses
of a plurality of circumferentially distributed withdrawal sensors are
converted into winding signals, then supplied to an evaluation unit and
compared with an expected signal pattern which corresponds to a
predetermined time sequence of the winding signals during
interference-free operation. The winding signals are only taken into
account for the control of the weft-yarn store when the time sequence of
the supplied winding signals complies with the expected pattern.
Furthermore, prior use in practice has revealed a method in which each yarn
pulse is converted into a winding signal in a filter means assigned to the
receiver of the withdrawal sensor and in which the winding signal helps to
open a time window within which successive pulses or signals are ignored.
Such a measure prevents lint clusters following at a lower speed from
leading to winding signals within the time window. However, when two
closely adjacent windings are withdrawn, then the second winding can no
longer be detected, which leads to an excessively long pick.
In modern, fast air-jet weaving machines it sometimes happens for reasons
which are not exactly understood that, for instance, once every 1000
insertion cycles there is an insertion in the case of which the yarn is
inserted at a slower pace than has been predetermined. Such an insertion,
however, is not to effect a shut-off of the weaving machine, for the
insertion is per se correct, but only too slow. Furthermore, it has been
found in practice that with specific yarn qualities not only lint clusters
are separately entrained after the yarn windings, but yarn components
which are still clinging to the yarn are also dragged along, e.g., in the
case of multifilament yarns. Such lint clusters or clinging components
then produce other types of pulses (with flat ramp and low frequency
content) than the yarn itself. Such false yarn pulses caused by entrained
yarn components are also not meant to produce winding signals. By
contrast, two winding signals are actually to be produced when two
windings are simultaneously withdrawn, as is sometimes the case. These
above-mentioned circumstances create special requirements for the
withdrawal sensor which is to make a reliable distinction between yarn
windings and other objects.
It is the object of the present invention to provide a simple method of the
above-mentioned kind and a yarn withdrawal sensor for performing said
method with the help of which short and long picks are avoided when using
a measuring and feeding device a weaving machine.
SUMMARY OF THE INVENTION
This object is achieved with the yarn withdrawal sensor and method therefor
wherein filter means such as a band-pass filter assembly is shiftable
between first and second filtering modes to prevent processing of
interference pulses.
The method prevents the production of wrong winding signals due to lint
clusters or other dirt moving at a pace slower than the yarn or producing
weak interference pulses, since the set yarn pulse acceptance for strong
and fast yarn pulses rules out a situation where the filtering means
accepts the slower or weak interference pulse of a lint cluster. A finding
which has been made in practice is here taken into account, i.e., the
finding that during the initial slow yarn withdrawal in the acceleration
phase of the insertion cycle there are hardly any passing lint clusters or
contaminations at any rate. Such interference is most of the time only
observed during fast yarn withdrawal after the acceleration phase. Thanks
to the division of each insertion process into at least one portion for
low yarn speed and into at least one portion for increased yarn speed, the
division being performed by intentionally varying the yarn pulse
acceptance, all sorts of yarn pulses are properly recorded, but no winding
signals are produced from interference pulses. If use were only made of an
unvaried broad yarn pulse acceptance equally suited for slow and fast yarn
pulses, no distinction could be made between slow and correct initial yarn
pulses and interference pulses at a higher yarn speed because the
interference pulses (caused by moving dirt) would lead to winding signals
during scanning as does(do) also the slow initial yarn pulse(s). A special
advantage is here that the method properly produces two winding signals
also in the case of two yarn windings passing shortly one after the other
through the withdrawal sensor without the yarn pulse of the second one of
said two yarn windings being lost. Since the winding signal information
for measuring the predetermined yarn length is reliable and unalterable,
weft yarns which are too short or too long are avoided despite unavoidable
dirt and despite the fact that some windings are sometimes withdrawn at
almost the same time. The steps of the method can also be carried out
manually in case of repair work or when a yarn feeding device is adjusted
for the first time or is run in. A strong or fast "pulse" means an
electrical signal which has a steep ramp and a high frequency portion. A
weak or slow "pulse" has no steep ramp and a small frequency portion.
In the yarn withdrawal sensor, the at least first yarn pulse and above all
the fast yarn pulses are distinguishable from slow or weak interference
pulses with the aid of the two different selective filtering modes. The
second selective filtering mode does not accept any interference pulses
which would be slower or weaker than the strong yarn pulses. As for the
interference pulses caused by dirt, it should be noted that lint clusters
have most of the time an increased extension in the direction of passage
below the withdrawal sensor and also a different appearance, so that the
interference pulse is detected to be a slower or weaker one not only
because of the low speed of movement of such a contamination, e.g. a lint
cluster, but also because of the increased extension and the other
characteristic, e.g. reduced density. An interference pulse caused by dirt
has a leading ramp which is not so steep and a frequency portion
(frequency content) smaller than that of the previous strong yarn pulse,
the leading ramp of the respective pulse being an important criterion for
the derivation of the winding signal. The respective acceptance or the
respective filtering mode is chosen such that "slow or weak" interference
pulses are filtered out.
BRIEF DESCRIPTION OF THE DRAWINGS
Embodiments of the subject matter of the invention will now be explained
with reference to the drawings, in which:
FIG. 1 is a diagrammatic side view of a weaving machine including a
measuring and feeding device;
FIG. 2 is a diagrammatic top view onto the measuring and feeding device of
FIG. 1;
FIG. 3 shows an alternative embodiment of a measuring and feeding device;
FIGS. 4 and 5 schematically show a first type of method, wherein the upper
diagram represents yarn and interference pulses as well as winding signals
generated from the yarn pulses, whilst the lower diagram illustrates
frequency bandwidths which are assigned to specific yarn speed ranges;
FIGS. 4A and 5A show two diagrams of a second variant of the method,
corresponding to FIGS. 4 and 5;
FIG. 6 shows a schematic block diagram of a simplified embodiment of a
circuit of a withdrawal sensor;
FIG. 7 shows a detailed block diagram of an electric circuit of a
withdrawal sensor; and
FIGS. 8a-8h show the response characteristics of the withdrawal sensor.
DETAILED DESCRIPTION
FIG. 1 diagrammatically illustrates a weft-yarn feeding device F (measuring
and feeding device) of known construction which supplies a weft yarn Y
intermittently and at a respectively exactly dimensioned identical length
to a shed H of a weaving machine L, e.g., a jet loom. Yarn Y is withdrawn
from a supply coil (not shown), it is passed through a motor housing 1 and
wound in a winding reservoir 3 onto a storage drum 2 from which it is
withdrawn below a stop device 4 and overhead by means of an insertion
device 6, e.g., an insertion nozzle. The stop device 4 is oriented with a
stop element towards a withdrawal region of the storage drum 2. Stop
element 5 is activated and deactivated for proportioning the yarn length
by means of a control device C. In the activated state of stop element 5,
yarn Y is retained. In the deactivated state of stop element 5, yarn Y can
freely be unwound from the winding reservoir 3. Winding reservoir 3 is
replenished with the help of a drive (not shown) of the feeding device F
in the customary manner so that new yarn is wound up. With the aid of a
withdrawal sensor S, a winding signal is produced during each passage of
yarn Y through a scanning portion positioned below the withdrawal sensor
S. The winding signals will be counted until the predetermined yarn length
has been reached. The stop element 5 is then activated again. The
circumferential length of the storage drum may be variable for adjusting a
predetermined yarn length for all insertion cycles.
FIG. 2 is a top view on the feeding device F, in which the withdrawal
sensor S is arranged in the motional direction of yarn Y just behind the
stop element 5 during withdrawal (arrow), expediently offset in the axial
direction of storage drum 2 relative to the stop element 5, e.g. by about
1 cm, to ensure a relatively high passage speed of yarn Y at a receiver R
of the withdrawal sensor S at all times. Yarn Y extends from the last
winding of the winding reservoir 3 in a direction oblique to stop device
4, it is deflected at stop element 5 in the activated state of said stop
element 5 and extends away from stop element 5 at the widthdrawal side
approximately in axial direction.
In modern weaving machines L, yarn speeds of up to 100 m/s or more are
observed during insertion. However, the yarn Y must first be accelerated
to reach its maximum yarn speed after each deactivation of stop device 4,
5. This means that after the deactivation of stop element 5 yarn Y passes
through the withdrawal sensor S at a relatively low speed of, for
instance, slightly more than 2 m/s--at least the first time, but already
runs at a substantially higher speed (in the direction of the arrow)
during its next passage underneath the withdrawal sensor S. With most yarn
qualities, contaminations, such as lint clusters which are entrained
during unwinding of the yarn and possibly also pass through the scanning
portion below the withdrawal sensor S, can be in yarn reservoir 3. Such
dirt, however, normally moves at a slower pace than the yarn and it is
recorded by the withdrawal sensor more slowly than the yarn. Yarn
components which extend away from the yarn may also be entrained; such
components, however, also produce weak interference pulses.
FIG. 3 shows a modified embodiment of a feeding device F in which a
withdrawal sensor S is respectively arranged at both sides of the stop
device 4 at a short distance. The two withdrawal sensors S form, for
instance, one winding signal based on the two yarn passages. Thanks to the
two withdrawal sensors S, the feeding device F can selectively be operated
in the one or the other rotational direction. It is possible to then use
only one of the two withdrawal sensors S.
In FIG. 4, the upper diagram part illustrates (electric) yarn pulses
effected by the yarn Y in the withdrawal sensor S of FIGS. 1 and 2 during
an insertion cycle. After stop element 5 has been deactivated, a first
slow and weak yarn pulse YP1 is first produced, the speed of said yarn
pulse YP1 being respresented by the width and a relatively flat leading
ramp. The next occurring yarn pulses YP2 are faster and stronger, which is
expressed by their steeper leading ramp (higher frequency portion or
content) and their pointed shape. Interference pulses LP1 and LP2, which
are outlined in broken line, are caused by dirt or by yarn components that
are possibly torn away during withdrawal and pass through the scanning
portion behind the yarn. The first interference pulse LP1 is slower and
weaker than the first yarn pulse YP1, however, said interference pulse LP1
being very unlikely for the reason that, at the beginning of the
withdrawal process, dirt is hardly entrained because of the slow
withdrawal speed of the yarn and because of the absence of any pronounced
air turbulences or dynamics. Normally, such an effect is only observed at
an increased yarn speed. The further interference pulses LP2 are slower or
weaker than the faster yarn pulses YP2. It may also happen, particularly
in the dynamic phase during an insertion cycle, that two yarn windings are
removed from the winding reservoir 3 and are unwound almost
simultaneously, i.e., in close vicinity next to each other. Such a
situation is demonstrated in FIG. 4 by the second faster yarn pulse YP2
and the directly following faster yarn pulse YP2'.
The lower diagram part of FIG. 4 shows how winding signals WP are produced
for the control device C from the yarn pulses YP1, YP2, YP2'. A winding
signal WP which is representative of the yarn passage is produced on the
basis of the first yarn pulse YP1. As soon as, for instance, the first
winding signal WP has been recorded, the withdrawal sensor S is switched
over or adjusted such that it only produces winding signals WP on the
basis of faster and strong yarn pulses YP2, YP2'. The adjustment is made
at time X. After the adjusting process the withdrawal sensor does not
produce any winding signals WP from the slower or weaker interference
pulses LP2. False winding signals caused by such interference pulses are
thereby avoided. By contrast, two winding signals WP and WP'are produced
in a proper manner even when two neighboring windings (two strong yarn
pulses YP2, YP2') are unwound almost simultaneously.
Each yarn pulse is electrically produced and processed in an electric
filter assembly E (FIGS. 6 and 7). The filter assembly includes, for
instance, bandpass filters the frequency bands of which are outlined in
FIG. 5. In the first set mode of the withdrawal sensor, the filter
assembly operates with a frequency range f1 between a lower frequency
limit fU1 and an upper frequency limit fO. fU1 corresponds, for instance,
to a minimum yarn speed of 2 m/s. fO corresponds, for instance, to a speed
of 120 m/s of the yarn (Vmax). At time X the filter assembly is upshifted
to a different frequency range f2 whose lower limit fU2 is higher than the
lower limit fU1. fU2 corresponds, for instance, to a minimum yarn speed of
10 m/s. In the second set mode after time X the upper limit fO is the same
as before.
In the set mode of the withdrawal sensor before time X, the at least first
slow and weak yarn pulse YP1 is just about accepted. After time X fast and
strong yarn pulses YP2, YP2'are accepted, whereas slow or weak
interference pulses LP2 are not accepted.
Upon deactivation of stop element 5, the first frequency range f1 is set.
After generation of the first winding signal WP, there is switching over
to the second frequency range f2 either with the help of this winding
signal or possibly with the second winding signal, or in response to the
known increase in yarn speed at time X. When the stop device is again
activated, the filter assembly is again reset into the first range f1.
For repair work or adjustments or for running in the yarn feeding device,
the withdrawal sensor may also be switched over by hand, the automatic
position which is employed during normal operation of the yarn feeding
device being then neutralized.
The diagrams of FIGS. 4A and 5A represent variants of the method in which
the filtering mode f1 is adjusted prior to the generation of each winding
signal WP, and in which upon generation of each winding signal WP there is
switching over to the further filtering mode f2. An upshifting signal
which is maintained over a time window H whose duration t.sub.F is
uniformly predetermined, e.g. at 3 ms, is respectively produced at time X.
The time window H is opened by means of the counting or timing member Z of
FIG. 7, for instance by each winding signal WP, and upon output of the
upshifting signal Fl at time X. After the time window H has expired, there
is again downshifting to filtering mode f1. If there is switching between
the filtering modes during the whole insertion cycle, detection errors are
also avoided during a slow insertion that is not observed very often. The
time window H is outlined in FIG. 5A only diagrammatically and not true to
scale. It should extend over a period of time within which the passage of
dirt must be expected.
FIGS. 8a-8h are concrete function diagrams of the withdrawal sensor S.
The diagrams (d.c. voltage across the logarithmically represented
frequency) of FIGS. 8a and 8b represent the response characteristics of
the band-pass filter assembly to yarn pulses. In FIG. 8a, a high-pass
filter mode and a low-pass filter mode are operative at the same time. The
filter assembly has a spread response range in which frequencies which are
clearly below 1.0 kHz already yield a d.c. voltage level of 0.6 V or more,
a d.c. level of about 0.8 V is obtained over a frequency range of from 1.0
kHz to about 20 kHz, and a d.c. voltage level of about 0.6 V is even
obtained at a frequency of 100 kHz.
In FIG. 8b the low-pass filter mode has been disabled, so that the response
characteristics in the diagram of the d.c. voltage across the frequency in
the upper frequency range remain approximately the same as in FIG. 8a, but
are different in the lower frequency range. A frequency which is clearly
below 10 kHz just yields a d.c. voltage of 0.6 V, frequencies between 10
kHz and 70 kHz yield d.c. voltage levels of 0.8 V and more, and
frequencies of 100 Hz to about 7.0 kHz yield d.c. levels which are clearly
below 0.6 V.
FIG. 8c illustrates the input signal to the band-pass filter means in the
form of a d.c. level curve over time (ms) during the first yarn passage,
the input signal extending up to a d.c. voltage value of about -1.0 V and
lasting for about 0.5 ms. The associated diagram of FIG. 8d represents the
associated output signal of the band-pass filter assembly after response
to the signal shown in the diagram of FIG. 8c. As can be seen, a strong
output signal with an absolute d.c. voltage value of almost 2.0 V over
approximately 0.5 ms is observed in the response characteristics according
to FIG. 8a (low-pass filter mode and high-pass filter mode).
FIGS. 8e and 8f are diagrams (d.c. voltage over time) which represent the
input signal to and the output signal from the band-pass filter assembly,
i.e. in the case of response characteristics according to FIG. 8b
(low-pass filter mode disabled) and with the same input signal as in FIG.
8c, i.e., during passage of dirt with a pulse format corresponding to a
yarn pulse format. Since the signal curve in FIG. 8e contains only weak or
few frequency portions due to reduced steepness of its leading or trailing
edge, respectively, a level of less than 0.1 V is just obtained as the
output signal of the band-pass filter means in FIG. 8f, which is ignored
and does not lead to a useful signal.
FIGS. 8g and 8h represent the response of the band-pass filter assembly to
a faster yarn pulse YP2 which is shown in the diagram of FIG. 8g (voltage
over time) as a strong signal of up to -1.0 V over a period of about 0.1
ms and with a virtually vertical drop and vertical rise, i.e. a high
frequency portion. This is the input signal of the band-pass filter
assembly from which the output signal of FIG. 8h is produced in the
band-pass filter assembly, which output signal is obtained as a distinct
winding signal WP with a voltage level of about 1.0 V and a subsequent
drop to almost -1.0 V and which can clearly be distinguished from the
considerably weaker signal of FIG. 8f caused by an interference pulse LP2.
FIG. 6 diagrammatically illustrates an embodiment of a circuit D of the
withdrawal sensor S between a receiver R and control unit C or a
microprocessor MP. The yarn pulse produced by receiver R is supplied to an
operational amplifier 7 behind which, in this embodiment, two band-pass
filters 8a and 8b are arranged in parallel, said band-pass filters 8a and
8b having respectively arranged downstream thereof members 9a, 9b for
producing the winding signals. The two band-pass filters 8a and 8b have
different frequency ranges f1, f2. A switching device 10 is connected via
a line 11 to control C and microprocessor MP, respectively, and is
switchable between two switching positions to activate either the one
branch or the other branch of the filter assembly. Upshifting (and
resetting) is performed by way of an upshifting signal (and resetting
signal, respectively) from the control device or microprocessor C, MP,
i.e., either upon generation of the at least first winding signal or in
response to the yarn speed which is normally measured, i.e. when a
predetermined yarn speed is reached that is representative of the yarn
having passed through the withdrawal sensor for the first time, or with
each winding signal (FIGS. 4A, 5A).
FIG. 7 shows a circuit having a band-pass filter assembly E and a
sensitivity adjusting device G with the aid of which the withdrawal sensor
S can be adapted to the respective yarn quality and working conditions.
Receiver R is connected to a positive input 27 of an operational amplifier
12 from the output 29 of which a feedback loop 30 leads to the negative
input 28 thereof. A resistor R21 is accommodated in the feedback loop 30.
A terminal 31 of an analog circuit component 12 which is virtually
grounded at Vvg is connected between resistor R21 and the negative input
28 via a resistor R22. At 32, a sensitivity adjusting signal AMP can be
applied to the analog circuit component 12, for instance a higher or a
lower voltage level (digital 1 or 0) which is provided via a line 22 by
microprocessor MP.
A capacitor C14 and a resistor R17 behind which a virtually grounded
junction 33 is provided are arranged downstream of the output 29 of the
operational amplifier 12. Capacitors and resistors C12, R5 and C12, R4,
R18, R5 are provided in parallel behind said junction 33. Capacitor C12 is
directly connected to a winding signal output 20, and, in addition, via
resistor R5 to an output 17 of a further operational amplifier 16. The
input of capacitor C13 is connected via resistor R18 to a terminal 25 of a
further analog circuit component 14 which is virtually grounded and has a
terminal 26 to which an upshifting signal Fl can be applied, typically a
voltage level which is provided via a line 21 by microprocessor MP, e.g.
upon receipt of the first or a respective winding signal WP.
The output of capacitor C13 is also connected to the negative input 17 of
the operational amplifier 16 whose output 19 is connected to the winding
signal output 20. The positive input 18 of the operational amplifier 16 is
virtually grounded (VVg). A further analog circuit component 15 is
arranged between the negative input 17 of the operational amplifier 16 and
the line which extends from capacitor C12 to the winding signal output 20,
a resistor R4 being used between the terminal 22 thereof and the negative
input 17. A terminal 24 of the analog circuit component 15 can be fed with
the upshifting signal Fl which is provided via a line 21 by microprocessor
20. The microprocessor MP may have a timing or counting member Z to
maintain the upshifting signal Fl via a predetermined period (t.sub.F in
FIG. 5A) upon generation of the winding signal WP, for instance over 3 ms.
The period t.sub.F is shorter than the time interval between two winding
signals WP at maximum withdrawal speed (e.g. 10 ms), preferably and for
reasons of safety even shorter than half this time interval.
The sensitivity adjusting signal AMP is either a low or a high voltage
level. Similarly, the upshifting signal Fl is produced as a high voltage
level (digital 1 or 0).
In FIG. 7, no upshifting signal Fl is present at inputs 24 and 26 of the
analog circuit components 14, 15 (i.e. a digital "0"). Hence, the
frequency range f1 has been selected and the low-pass filtering mode has
been enabled. In response to the yarn quality and the working conditions,
there is either a digital 1 or a digital 0 present as the sensitivity
adjusting signal AMP. A winding signal which is received by the
microprocessor MP is produced from the at least first yarn pulse.
Thereupon, a "digital 1" is produced as the upshifting signal Fl. The
circuit is switched over to the second frequency range f2 (low-pass
filtering mode disabled), whereby the anlog circuit components 14, 15
change the resistance characteristics of resistors R4, R18. To prevent
such a change from influencing the sensitivity adjustment, the circuit
components 13, 15, 14 are grounded and junction 33 is also grounded to
ensure that the respective d.c. level will not drift off as soon as there
is a switching operation. An impact on the sensitivity adjusting device G
is thereby avoided. Such an impact is mainly operative by changing the
amplifier factor in the operational amplifier 12. When a digital "0" is
present as the sensitivity adjusting signal AMP, the amplification factor
is, for instance, "1". When a digital "1" is present as the AMP, the
amplification factor is 1+R21:R22. When the yarn is stopped after
insertion or when the microprocessor MP no longer receives a winding
signal over a long period of time or when the time window H in FIG. 4A, 5A
has expired, the circuit D is reset via line 21 to the adjustment of the
first frequency range f1.
The withdrawal sensor S is not necessarily arranged in the same radial
plane as the stop device. The withdrawal sensor S could also be arranged
in axial direction at the side of the stop device which is oriented away
from the winding reservoir, i.e. also in front of the face of storage drum
2.
As discussed herein, the band-pass filtering means switches from a first
filtering mode to a second filtering mode as soon as the first yarn pulse
has led to the first winding signal. The subsequent yarn pulses are then
so fast and strong that they are accepted in the second filtering mode
whereas "slow or weak" interference pulses are not accepted.
In a particularly expedient variant of the method, prior to each arising
yarn pulse or winding signal, the yarn pulse acceptance is temporarily
adjusted for a weak yarn pulse before the yarn pulse acceptance is again
changed subsequently with the winding signal to the yarn pulse acceptance
for faster and strong yarn pulses. Thanks to the yarn pulse acceptance for
faster and strong yarn pulses, which is only set for a short period of
time, slow or weak interference pulses are prevented from producing a
winding signal. A further effect is that also in the case of a slow
insertion, which does not happen very often, the yarn pulses reliably lead
to winding signals because before each yarn pulse the yarn acceptance is
adjusted for weak yarn pulses, and a contamination following the yarn does
not lead to a wrong winding signal because in such a case the yarn
acceptance is then set for strong winding signals.
In a further variant, the time within which a weak interference pulse uses
to occur is, so to speak, cut out in a technical control process simply by
means of the time window. The time window is, for instance, set to 3 ms,
i.e., a duration which is shorter than the shortest time interval between
two successive winding signals of the insertion cycle, typically at least
10 ms.
Another variant is simple and reliable with respect to control. Upon
detection of the at least first winding signal, the upshifting signal is
supplied to the filter assembly so that the assembly operates with the
second filtering mode for the further yarn pulses. The upshifting signal
may be produced with a time delay. Downshifting into the first filtering
mode is carried out after expiration of the time window and prior to the
occurrence of the next yarn pulse by means of suppressing the upshifting
signal.
The filtering means is a band-pass filter assembly having two different
bandwiths, with the lower limit of a bandwith being respectively adjusted
to the rapidity of the yarn pulse and to the slowness or weakness of the
interference pulses to be able to make a distinction between the two.
Where the band-pass filter assembly is connected to a microprocessor, the
band-pass filtering means is switched over by the microprocessor as soon
as the at least first or each winding signal has been produced.
In another embodiment, the filter assembly is permanently switched back and
forth between the two filtering modes in such a manner that prior to the
generation of a yarn pulse the filtering mode is set with acceptance of
also a slow or weak yarn pulse, whereas upon generation of the winding
signal and for the duration of the time window the filtering mode remains
set with acceptance of only strong yarn pulses. Hence, interference pulses
can be filtered out, and above all a very rarely occurring slow insertion
into the weaving machine can be mastered without slow insertion into the
weaving machine can be mastered without any detection errors.
The embodiment wherein the control circuit includes an active amplifier and
band-pass filter assembly is particularly expedient because the active
amplifier and band-pass filter assembly lead to uniformly strong and
significant winding signals and avoids performance losses during
filtering.
The band-pass filter assembly is designed with response characteristics
including a high-pass filtering mode and a low-pass filtering mode
following said high-pass filtering mode without any interruption. A
significant d.c. level which up to higher frequencies remains
approximately constant around about 100 kHz is thereby obtained up to
frequencies of, for instance, less than 1.0 kHz. The low-pass filtering
mode can be disabled to change the response characteristics such that
frequencies of, for instance, clearly less than 10 kHz or a frequency of
about 1.0 kHz no longer lead to any significant d.c. level, but only
frequencies of between about 10 khZ and shortly below 100 kHz lead to
similarly high or higher d.c. levels as in the effective low-pass
filtering mode. This can easily be achieved with a control means by
varying the resistance characteristics of the two resistors; of special
importance is here the fact that analog circuit components to which the
two resistors are connected ensure that the d.c. level in the band-pass
filter assembly is kept constant and does not drift off because of the
switching between the two modes. In other words, the band-pass filter
assembly has response characteristics which first lead via a relatively
broad frequency range to a significant d.c. level, but, in case of need,
are temporarily restricted by disabling the low-pass filtering mode to a
narrow frequency range near the upper cut-off frequency, so that only
higher frequencies lead to usable d.c. levels. In a disabled low-pass
filtering mode, the interference pulses at lower frequencies can thus be
filtered out at lower frequencies, because only the yarn pulses with the
correspondingly high frequencies lead to high d.c. levels.
The bandwiths are dimensioned such that the yarn speeds which are normally
high in modern weaving machines can be mastered without any difficulty.
Further, it is ensured that the band-pass filter assembly respectively
operates at the beginning of the insertion process and before each yarn
pulse in the first filtering mode.
In an alternative simple embodiment, there is switching between the
band-pass filters, depending on whether the yarn moves at a low speed or
at a high speed or whether or not the yarn has passed the withdrawal
sensor.
Further, the withdrawal sensor serves to control the stop device to exactly
dimension the yarn length. The receiver is positioned closely behind the
stop device to report the proper passage of the yarn as early as possible.
Also, the receiver is offset in the axial direction of the storage drum
relative to the stop element of the stop device, expediently at the side
of the stop element which faces the yarn reservoir, in order to obtain a
yarn geometry with an obliquely extending yarn in the activated state of
the stop, the yarn geometry being such that a specific amount of time will
elapse upon deactivation of the stop element until the yarn has passed the
receiver. Thanks to this elapsed time and on account of the strong
acceleration at the beginning of the yarn insertion cycle, the passage
speed at the receiver will then be already so high that a relatively
strong first yarn pulse is generated.
By contrast, the two alternately or jointly activated receivers are
provided at both sides of the stop device in order to achieve an even
higher accuracy during scanning. Each winding signal is derived from two
successive yarn pulses. This arrangement allows changing of the rotational
withdrawal direction as well.
The withdrawal sensor can be adapted to the respective yarn quality, with
an undesired interaction between the change in acceptance or the switching
between the filtering modes and the sensitivity adjustment being avoided
by uncoupling. The sensitivity must be adjusted because different yarn
qualities may lead to different yarn pulses, e.g. because of different
reflective characteristics or densities.
The withdrawal sensor expediently operates in an optoelectric manner.
However, it is also possible to scan the yarn without contact by
ultrasound, in a capacitive, inductive or piezoelectric manner, or with
contact. A precondition is that the receiver is capable of producing yarn
pulses having a specific pulse shape or a specific curve of the leading
ramp.
Top